PAC-Bayesian Bounds for Randomized Empirical Risk Minimizers - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Mathematical Methods of Statistics Année : 2008

PAC-Bayesian Bounds for Randomized Empirical Risk Minimizers

Résumé

The aim of this paper is to generalize the PAC-Bayesian theorems proved by Catoni in the classification setting to more general problems of statistical inference. We show how to control the deviations of the risk of randomized estimators. A particular attention is paid to randomized estimators drawn in a small neighborhood of classical estimators, whose study leads to control the risk of the latter. These results allow to bound the risk of very general estimation procedures, as well as to perform model selection.
Fichier principal
Vignette du fichier
pacbayesMMS.pdf (336.92 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-00195698 , version 1 (11-12-2007)
hal-00195698 , version 2 (04-11-2008)
hal-00195698 , version 3 (09-01-2009)

Identifiants

Citer

Pierre Alquier. PAC-Bayesian Bounds for Randomized Empirical Risk Minimizers. Mathematical Methods of Statistics, 2008, 17 (4), pp.279-304. ⟨10.3103/S1066530708040017⟩. ⟨hal-00195698v3⟩
304 Consultations
184 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More